Alterations in the structural and functional states of chromatin, mainly determined by post-translational modifications of histone components, are involved in the pathogenesis of a variety of diseases. The enzymatic processes, which govern these post-translational modifications on the nucleosomes, have become potential targets for the so-called epigenetic therapies (Portela, A. et al. Nat. Biotechnol. 2010, 28, 1057-1068). The discovery of an increasing number of histone lysine demethylases has highlighted the dynamic nature of the regulation of histone methylation, a key chromatin modification that is involved in eukaryotic genome and gene regulation. Histone lysine demethylases represent attractive targets for epigenetic drugs, since their expression and/or activities are often misregulated in cancer (Varier, R. A. et al. Biochim. Biophys. Acta. 2011, 1815, 75-89). A lysine can be mono-, di-, and tri-methylated and each modification, even on the same amino acid, can have different biological effects.
Histone lysine demethylases exert their activity through two different type of mechanism (Anand, R. et al. J. Biol. Chem. 2007, 282, 35425-35429; Metzger, E. et al. Nat. Struct. Mol. Biol. 2007, 14, 252-254). While the Jumonji domain-containing histone demethylases, which are iron and 2-oxoglutarate dependent oxygenases, act on mono-, di- and trimethylated lysines, the flavin-dependent (FAD) histone demethylases catalyse the cleavage of mono and dimethylated lysine residues. Currently, two FAD dependent demethylases have been identified: LSD1, also known as KDM1A, and LSD2, also known as KDM1B. (Culhane, J. C. et al. Curr. Opin. Chem. Biol. 2007, 11, 561-568, Ciccone, D. N. et al. Nature 2009, 461, 415-418).
KDM1A is a constituent in several chromatin-remodeling complexes and is often associated with the co-repressor protein CoREST. KDM1A specifically removes the methyl groups from mono- and di-methyl Lys4 of histone H3, which is a well-characterized gene activation mark. KDM1A represents an interesting target for epigenetic drugs due to its over-expression in solid and hematological tumors (Schulte, J. H. et al. Cancer Res. 2009, 69, 2065-2071; Lim, S. et al. Carcinogenesis 2010, 31, 512-520; Hayami, S. et al. Int. J. Cancer 2011, 128, 574-586; Schildhaus, H. U. et al. Hum. Pathol. 2011, 42, 1667-1675; Bennani-Baiti, I. M. et al. Hum. Pathol. 2012, 43, 1300-1307). Its over-expression correlates to tumor recurrence in prostate cancer (Kahl, P. et al. Cancer Res. 2006, 66, 11341-11347), and KDM1A has a role in various differentiation processes, such as adipogenesis (Musri, M. M. et al. J. Biol. Chem. 2010, 285, 30034-30041), muscle skeletal differentiation (Choi, J. et al. Biochem. Biophys. Res. Commun. 2010, 401, 327-332), and hematopoiesis (Hu, X. et al. Proc. Natl. Acad. Sci. USA 2009, 106, 10141-10146; Li, Y. et al. Oncogene. 2012, 31, 5007-5018). KDM1A is further involved in the regulation of cellular energy expenditure (Hino S. Et al. Nat Commun. 2012, doi: 10.1038/ncomms1755), in the regulation of thermogenesis and oxidative metabolism in adipose tissue (Duteil et al. Nat Commun. 2014 Jun. 10; 5:4093. doi: 10.1038/ncomms5093.), in the control of checkpoints of viral gene expression in productive and latent infections (Roizman, B. J. Virol. 2011, 85, 7474-7482), and more specifically in the control of herpes virus infection (Gu, H. J. Virol. 2009, 83, 4376-4385) and HIV transcription (Sakane, N. et al. PLoS Pathog. 2011, 7(8):e1002184). The role of KDM1A in the regulation of neural stem cell proliferation (Sun, G. et al. Mol. Cell Biol. 2010, 30, 1997-2005) and in the control of neuritis morphogenesis (Zibetti, C. et al. J. Neurosci. 2010, 30, 2521-2532) suggests its possible involvement in neurodegenerative diseases.
Furthermore, KDM1A has been found to be relevant in the control of other important cellular processes, such as DNA methylation (Wang, J. et al. Nat. Genet. 2009, 41(1):125-129), cell proliferation (Scoumanne, A. et al. J. Biol. Chem. 2007, 282, 15471-15475; Cho, H. S. et al. Cancer Res. 2011, 71, 655-660), epithelial mesenchimal transition (Lin, T. et al. Oncogene. 2010, 29, 4896-4904) and chromosome segregation (Lv, S. et al. Eur. J. Cell Biol. 2010, 89, 557-563). Moreover, KDM1A inhibitors were able to reactivate silenced tumor suppressor genes (Huang, Y. et al. Proc. Natl. Acad. Sci. USA. 2007, 104, 8023-8028; Huang, Y. et al. Clin. Cancer Res. 2009, 15, 7217-7228), to target selectively cancer cells with pluripotent stem cell properties (Wang, J. et al. Cancer Res. 2011, 71, 7238-7249), as well as to reactivate the all-trans-retinoic acid differentiation pathway in acute myeloid leukemia (Schenk, T. et al. Nat Med. 2012, 18, 605-611). Moreover, KDM1A has a clear role in sustaining the oncogenic potential of MLL-AF9 translocation in leukaemia stem cells (Harris et al. Cancer Cell, 21 (2012), 473-487), as well as in the stem-like tumor propagating cells of human glioblastoma (Suvà et al. Cell 2014, 157, 580-594).
The more recently discovered demethylase KDM1B (LSD2) displays similarly to KDM1A specificity for mono- and di-methylated Lys4 of histone H3. KDM1B, differently from KDM1A, does not bind CoREST and it has not been found up to now in any of KDM1A-containing protein complexes (Karytinos, A. et al. J. Biol. Chem. 2009, 284, 17775-17782). On the contrary, KDM1B forms active complexes with euchromatic histone methyltransferases G9a and NSD3 as well as with cellular factors involved in transcription elongation. KDM1B has been reported to have a role as regulator of transcription elongation rather than that of a transcriptional repressor as proposed for KDM1A (Fang, R. et al. Mol. Cell 2010, 39, 222-233).
KDM1A and KDM1B are both flavo amino oxidase dependent proteins sharing a FAD coenzyme-binding motif, a SWIRM domain and an amine oxidase domain, all of which are integral to the enzymatic activity of KDM1 family members. Moreover, both KDM1A and KDM1B show a structural similarity with the monoamine oxidases MAO-A and MAO-B. Indeed, tranylcypromine, a MAO inhibitor used as antidepressant agent, was found to be also able to inhibit KDM1A. The compound acts as an irreversible inhibitor forming a covalent adduct with the FAD cofactor. (Lee, M. G. et al. Chem. Biol. 2006, 13, 563; Schmidt, D. M. Z. et al. Biochemistry 2007, 46, 4408). Tranylcypromine analogs and their KDM1A inhibitory activity have been described in Bioorg. Med. Chem. Lett. 2008, 18, 3047-3051, in Bioorg. Med. Chem. 2011, 19, 3709-3716, and in J. Am. Chem. Soc 2011, 132, 6827-6833. Further arylcyclopropylamine and heteroarylcyclopropylamine derivatives as KDM1A, MAO-A and/or MAO-B enzyme inhibitors are disclosed in US2010/324147, in WO2012/045883, in WO2013/022047 and in WO2011/131576. Reversible KDM1A inhibitors are not so common and no clinical data for them are so far available. Examples of reversible inhibitors are aminothiazoles as described in Med. Chem. Commun. 2013, 4, 1513-1522, a N′-(1-phenylethylidene)-benzohydrazide series (J. Med. Chem. 2013, 56, 9496-9508, WO2013025805), or thienopyrrole derivatives (WO2016/034946). Thus, there is still a need for further reversible inhibitors having useful antitumor properties, adequate selectivity and stability of action, and possibly showing a higher activity with respect to specific subclasses thereof.